Using some real world production, energy and temperature data explore how a generative AI can be used to develop predictive models. In this case using ChatGPT 4o and o1. Interesting stuff. We go over some basic regression and tree models and ARIMA vs Prophet and allow the LLM to evaluate and assist us. Not how I really do it, but it shows the capabilities and weaknesses of generative and predictive AI (otherwise known as machine learning) and it allows us to explore the need for human interaction. Next week we will delve deeper into how it is done for production purposes.
Posted inArtificial Intelligence Data Science energy efficiency ESG Internet of Things Machine Learning MotorDoc Podcast Power Power Quality Uncategorized